Amazon's Profits Are Floating On a Cloud (Computing)

HughPickens.com writes: The NY Times reports that Amazon unveiled the financial performance of its powerful growth engine for the first time on Thursday, and the numbers looked good, energized primarily by renting processing power to start-ups and, increasingly, established businesses. Amazon said in its first-quarter earnings report that its cloud division, Amazon Web Services, had revenue of $1.57 billion during the first three months of the year. Even though the company often reports losses, the cloud business is generating substantial profits. The company said its operating income from AWS was $265 million.

Amazon helped popularize the field starting in 2006 and largely had commercial cloud computing to itself for years, an enormous advantage in an industry where rivals usually watch one another closely. At the moment, there is no contest: Amazon is dominant and might even be extending its lead. Microsoft ranks a distant No. 2 in cloud computing but hopes to pick up the slack with infrastructure-related services it sells through Azure, the name of its cloud service. Amazon executives have said they expect AWS to eventually rival the company's other businesses in size. The cloud business has been growing at about 40 percent a year, more than twice the rate of the overall company and many Wall Street analysts have been hoping for a spinoff.

As for Google, the cloud was barely mentioned in Google's earnings call. Nor did the search giant offer any cloud numbers, making it impossible to gauge how well it is doing. But the enthusiasm of Eric Schmidt, Google's executive chairman, was manifest when he spoke at an event for cloud software developers this week. "The entire world will be defined by smartphones, Android or Apple, a very fast network, and cloud computing," said Schmidt. "The space is very large, very vast, and no one is covering all of it."

Comments

Amazon has really been a stealth company

[surfdaddy]

They started as a bookseller, then moved slowly into other merchandise. Now they are eating the lunch of all sorts of brick and mortar stores. Then they went into cloud services and are grabbing that market. And now they are producing their own entertainment a la Netflix. They are a force to be reckoned with.

What is probably the saddest is Microsoft. Their two biggest cash cows (Windows and Office) are under tremendous pressure. And they really have trouble innovating in ways that are replacing that income. They chase everybody else, late to the game: mp3 players, search, cloud services, online email, smartphones, etc. Their constant focus on Windows over the Ballmer years really blinded them to all else that was an opportunity in the computing world. And so companies like Amazon, Apple, and Google are there instead.

Personal Anecdote

[Anonymous Coward]

I'm a Google Fanboy so when I wanted to set up a cloud server for a webapp I went straight to Cloud Compute. I had been burned by AWS phantom charges in the hundreds of dollars that could not be identified from their management console. Starcluster based OpenMPI "hello world" experiment turned $500 CC expense.

Anyway, right as I'm getting settled in to Cloud Compute(very happy with their tutorial on setting up mongoDB etc.) I'm building my web app and Cloud Compute suspends my account for terms of use violation out of the blue. No explanation. Just got flagged by heuristic analysis based on CC country vs. IP address continent probably(on vacation).

Whatever reason they flagged the account, I went through a nightmare ~1 week trying to get my account unfrozen. I eventually gave up and started a new AWS account which is now billing me $30/month for a Windows Server 2008 "Workspaces" instance I NEVER use. Moral: Google TOU "Kangaroo Courts" and shit customer support in appeals lost them my business, so I have a hard time crying for my favorite company on this front.

If they can't handle their shit with someone who WANTS to give them money I have no sympathy.

It is a cycle.

[140Mandak262Jamuna]

Back when IBM executive predicted "the world will probably need six computers", the main computing model was a mainframe at a distant location and time share on it via (overpriced) telephone lines and VT-100 terminals. Eventually workstations appeared and the move was to get off the mainframe and do local computing. Then came along Sun, "The network is *the* computer" and diskless workstations that would boot into an X-11 display terminal off a distant server. Well, PCs came along and desktop became powerful enough to run even fluid mechanics simulations. Then came high performance computing, and now the cloud.

A bigger machine in a far away place always had the cost advantages of the economy of scale. Everytime there is a jump in connection speeds and bandwidth some customers found it cheaper to "out source" computing to a remote machine. But eventually the advantages of local storage and local computation adds up. So let us see how long this iteration lasts.

Re:It is a cycle.

[fuzzyfuzzyfungus]

The one other element in the cycle you identify is arguably 'management/administration'.

This can work both for and against both local and remote/cloud options: Back when anything that touched the mainframe needed 6 signatures and a blessing; but you could classify an IBM-compatible as an 'office supply' and just have it on your desk and doing stuff, part of the virtue was in cutting through red tape, not in enjoying DOS on a slow machine with virtually no RAM. These days, especially for individuals or small outfits, without technical expertise available, 'the cloud' wins not so much because local computers are expensive(since they aren't, they've never been cheaper, either absolutely or per unit power); but because 'the cloud' is something you can use just by plugging in a URL and following directions. IT geeks are correct to point out that 'the cloud' is neither impregnable nor as well-backed-up as it likes to pretend to be; but for a non-techie user who will lose all their data as soon as their HDD dies or they lose their phone, it's still a step up.

For larger outfits, who have technical expertise available(and whose needs are complex enough that they will need IT and/or developers whether they go 'cloud', local, or some combination of the two), it is much more a straight battle on cost, security, and reliability; but ease of use and ease(or nonexistence) of management is huge for the consumer side.

Competition

[__aabppq7737]

Given that Microsoft seems to be investing heavily in Azure, I'd wonder exactly how they plan to beat AWS. AWS had some new machine learning algorithm added a month ago; Azure doesn't have that. Either way, however, is a win. If Microsoft's making some fatal mistake with their new business model, then maybe they'd go bankrupt and help the industry by going open-source before death. If Azure stays where it is or ranks up in usage with its SaaS model, then there'll probably be some interesting competition between them two and Google with large user bases. Either way, there's competition, which will (almost) forever spiral downward prices and upward capabilities.

Re:So, where's IBM in all of this?

[hawguy]

Seems like this should really be IBM's forte. I wonder why they didn't jump into it with both feet.

-jcr

Cheap commodity services was never IBM's forte - they don't want to rent you a $20 virtual server that you maintain yourself, they want to sell you a million dollar mainframe and $10,000 Intel servers that you pay IBM to maintain.